Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for generating a corrected image implemented on a magnetic resonance imaging (MRI) system including an MRI device and a computing device, the MRI device including multiple radio frequency (RF) receiver coils, the computing device including a processor, the method comprising: receiving, by the multiple RF receiver coils, MR data of an object; reconstructing, by the processor, a first MR image based on the MR data according to a first reconstruction algorithm; reconstructing, by the processor, a second MR image based on the MR data according to a second reconstruction algorithm, the second reconstruction algorithm being different with the first reconstruction algorithm; generating, by the processor, correction information about the first MR image by dividing the first MR image by the second MR image; and generating, by the processor, the corrected image with reduced inhomogeneity intensity based on the first MR image and the correction information about the first MR image.
This invention relates to magnetic resonance imaging (MRI) systems and addresses the problem of inhomogeneity artifacts in MR images, which can degrade image quality and diagnostic accuracy. The method involves using an MRI system with multiple radio frequency (RF) receiver coils and a computing device to process MR data and generate a corrected image with reduced inhomogeneity. The method begins by acquiring MR data of an object using the RF receiver coils. The computing device then reconstructs a first MR image from the MR data using a first reconstruction algorithm. Simultaneously, a second MR image is reconstructed from the same MR data using a second, different reconstruction algorithm. The processor then generates correction information by dividing the first MR image by the second MR image. This correction information is used to adjust the first MR image, producing a final corrected image with reduced inhomogeneity artifacts. The approach leverages the differences between the two reconstruction algorithms to identify and mitigate inhomogeneities in the MR data, improving image quality.
2. The method of claim 1 , the first reconstruction algorithm or the second reconstruction algorithm including at least one of a sum of squares (SOS) algorithm, a geometric average (GA) algorithm, a sensitivity encoding (SENSE) algorithm, a parallel imaging with localized sensitivities (PILS) algorithm, a modified sensitivity encoding (MSENSE) algorithm, or a SPACE RIP algorithm.
This invention relates to medical imaging, specifically magnetic resonance imaging (MRI) techniques for reconstructing images from acquired data. The problem addressed is improving image reconstruction quality and efficiency in MRI systems, particularly when using parallel imaging methods that rely on multiple receiver coils to accelerate scanning. The method involves using a first reconstruction algorithm and a second reconstruction algorithm to process MRI data. The first algorithm is applied to a first subset of the acquired data, while the second algorithm is applied to a second subset. The subsets may overlap or be distinct, and the algorithms may be applied sequentially or in parallel. The reconstruction results from both algorithms are then combined to produce a final image. This approach leverages the strengths of different reconstruction techniques to enhance image quality, reduce artifacts, and improve computational efficiency. The reconstruction algorithms may include various parallel imaging techniques such as sum of squares (SOS), geometric average (GA), sensitivity encoding (SENSE), parallel imaging with localized sensitivities (PILS), modified sensitivity encoding (MSENSE), or SPACE RIP. These algorithms are designed to handle the complex data acquired by multi-coil MRI systems, where each coil captures slightly different information. By combining results from multiple algorithms, the method aims to mitigate limitations inherent in any single approach, such as noise amplification or resolution loss. The technique is particularly useful in clinical settings where fast, high-quality imaging is required.
3. The method of claim 1 , the reconstructing an MR image based on the MR data including: generating multiple coil images based on the MR data, wherein each coil image is generated using a part of the MR data received by a corresponding RF receiver coil of the multiple RF receiver coils; and generating the MR image based on the multiple coil images.
Magnetic resonance imaging (MRI) systems use multiple radiofrequency (RF) receiver coils to acquire raw MR data, which is then processed to reconstruct high-quality images. A challenge in MRI is efficiently combining data from multiple coils to produce a final image while maintaining signal quality and reducing artifacts. This invention addresses this by improving the image reconstruction process. The method involves generating multiple coil images, each derived from a portion of the MR data received by a specific RF receiver coil. These coil images are then combined to form a single MR image. The approach ensures that data from each coil is properly utilized, enhancing image clarity and accuracy. By leveraging the distinct contributions of each coil, the method improves the overall image quality while mitigating noise and distortion. This technique is particularly useful in applications requiring high-resolution imaging, such as medical diagnostics, where precise and artifact-free images are critical. The method optimizes the reconstruction process, making it more efficient and reliable for clinical and research use.
4. The method of claim 3 , wherein the reconstructing a first MR image or the reconstructing a second MR image further includes: for each point of a plurality of points in the imaged object, determining pixel coordinates of corresponding pixels in the multiple coil images relating to the point of the imaged object; obtaining pixel values of the corresponding pixels in the multiple coil images of the point; and reconstructing the first MR image or the second MR image based on the pixel coordinates and the pixel values of the corresponding pixels in the multiple coil images of the plurality of points in the imaged object.
Magnetic resonance imaging (MRI) systems use multiple receiver coils to capture raw signal data from an imaged object. A challenge in MRI reconstruction is accurately combining signals from different coils to form a high-quality final image, especially when dealing with complex imaging scenarios. This invention addresses this problem by improving the reconstruction process for MRI images derived from multiple coil signals. The method involves reconstructing a first and/or second MR image from multiple coil images. For each point in the imaged object, the system determines the pixel coordinates of corresponding pixels in the multiple coil images. It then obtains the pixel values of these corresponding pixels across the coil images. The final MR image is reconstructed by processing these pixel coordinates and values from all relevant points in the object. This approach ensures that spatial and signal information from each coil is properly aligned and integrated, enhancing image accuracy and quality. The technique is particularly useful in applications requiring precise image reconstruction, such as dynamic imaging or high-resolution scans, where signal coherence across coils is critical. By systematically mapping and combining pixel data from multiple coils, the method improves the fidelity of the reconstructed MR images.
5. The method of claim 1 , wherein the correction information includes a divided image, and the generating the corrected image further includes: dividing the first MR image by the divided image.
This invention relates to medical imaging, specifically magnetic resonance (MR) imaging, and addresses the challenge of correcting distortions or artifacts in MR images to improve diagnostic accuracy. The method involves generating a corrected MR image by applying correction information derived from a divided image. The divided image is obtained by dividing a first MR image by another image, which may be a reference or calibration image. This division process helps isolate distortion patterns or artifacts present in the first MR image. The corrected image is then generated by further dividing the first MR image by this divided image, effectively normalizing the distortions and producing a cleaner, more accurate representation of the scanned anatomy. The technique is particularly useful in scenarios where MR images suffer from geometric distortions, intensity inhomogeneities, or other artifacts that degrade image quality. By leveraging the divided image as a correction factor, the method enhances the reliability of MR imaging for clinical diagnosis and treatment planning. The approach is adaptable to various MR imaging modalities and can be integrated into existing imaging workflows to improve image fidelity without requiring additional hardware.
6. The method of claim 1 , wherein the correction information includes a divided image, and the generating correction information further includes: smoothing the divided image to generate a smoothed divided image; and normalizing the smoothed divided image to generate a normalized image.
This invention relates to image processing techniques for correcting visual artifacts in digital images, particularly those caused by distortions such as lens aberrations, motion blur, or sensor noise. The method addresses the challenge of accurately restoring image quality by generating correction information that compensates for these distortions. The process involves dividing an input image into smaller segments to isolate and analyze specific distortion patterns. Each divided segment undergoes smoothing to reduce high-frequency noise and enhance structural coherence, resulting in a refined divided image. This smoothed image is then normalized to ensure consistent brightness and contrast across the segments, producing a normalized image that serves as correction information. The normalized image is applied to the original distorted image to correct distortions while preserving visual fidelity. The method ensures that the correction process is adaptive, handling varying distortion levels across different image regions. By smoothing and normalizing the divided segments, the technique minimizes artifacts and enhances overall image clarity. This approach is particularly useful in applications requiring high-precision image restoration, such as medical imaging, satellite photography, and high-resolution digital photography. The invention improves upon existing methods by providing a more localized and adaptive correction mechanism, leading to superior image quality.
7. The method of claim 6 , wherein the generating the corrected image further includes: dividing the first MR image by the normalized image.
This invention relates to medical imaging, specifically magnetic resonance (MR) imaging, and addresses the challenge of correcting distortions or artifacts in MR images to improve diagnostic accuracy. The method involves generating a corrected MR image by dividing a first MR image by a normalized image. The normalized image is derived from a second MR image, which is acquired with different imaging parameters to reduce artifacts. The second MR image is normalized by adjusting its intensity values to match the statistical distribution of the first MR image, ensuring consistency. The corrected image is then produced by dividing the first MR image by this normalized image, which helps mitigate intensity variations and artifacts caused by factors such as magnetic field inhomogeneities or patient motion. This process enhances image clarity and accuracy, making it useful for diagnostic and treatment planning purposes. The method is particularly valuable in applications where precise image quality is critical, such as brain imaging or tumor detection.
8. The method of claim 1 , wherein the multiple RF receiver coils have different spatial sensitivities and receive MR signals in parallel.
This invention relates to magnetic resonance imaging (MRI) systems, specifically improving signal acquisition by using multiple radiofrequency (RF) receiver coils with distinct spatial sensitivities. The problem addressed is the limited spatial resolution and signal-to-noise ratio (SNR) in conventional MRI systems, which rely on single or uniformly sensitive coils. The solution involves deploying multiple RF receiver coils, each designed to have unique spatial sensitivity profiles, allowing them to capture distinct portions of the imaging volume simultaneously. By receiving magnetic resonance (MR) signals in parallel, the system enhances data acquisition efficiency and image quality. The coils' differing sensitivities enable better localization of signals, reducing artifacts and improving contrast. This parallel acquisition approach also accelerates imaging by collecting complementary data streams, which can be combined to reconstruct high-resolution images faster than traditional methods. The invention is particularly useful in clinical and research settings where rapid, high-fidelity imaging is critical. The system may integrate with existing MRI hardware, requiring only modifications to the coil array and signal processing algorithms to handle parallel data streams. The method ensures that the combined signals from the coils are processed to form a coherent image, leveraging the spatial diversity of the coil array for superior performance.
9. A system comprising an MRI device and a computing device, the MRI device including multiple radio frequency (RF) receiver coils, the computing device including a processor, wherein during operation, the processor causes the system to: receive, by the multiple RF receiver coils, MR data of an object; reconstruct, by the processor, a first MR image based on the MR data according to a first reconstruction algorithm; reconstruct, by the processor, a second MR image based on the MR data according to a second reconstruction algorithm, the second reconstruction algorithm being different with the first reconstruction algorithm; generate, by the processor, correction information about the first MR image by dividing the first MR image by the second MR image; and generate, by the processor, a corrected image with reduced inhomogeneity intensity based on the first MR image and the correction information about the first MR image.
Magnetic resonance imaging (MRI) systems often suffer from inhomogeneity artifacts caused by variations in radio frequency (RF) coil sensitivity and other factors, leading to inconsistent image quality. This system addresses the problem by using multiple reconstruction algorithms to correct such inhomogeneities. The system includes an MRI device with multiple RF receiver coils and a computing device with a processor. During operation, the RF coils acquire MR data of an object. The processor then reconstructs a first MR image using a first reconstruction algorithm and a second MR image using a different second reconstruction algorithm. The processor generates correction information by dividing the first MR image by the second MR image, which helps identify and quantify inhomogeneities. Finally, the processor applies this correction information to the first MR image to produce a corrected image with reduced inhomogeneity intensity. This approach leverages multiple reconstruction techniques to improve image uniformity and accuracy.
10. The system of claim 9 , the first reconstruction algorithm or the second reconstruction algorithm including at least one of a sum of squares (SOS) algorithm, a geometric average (GA) algorithm, a sensitivity encoding (SENSE) algorithm, a parallel imaging with localized sensitivities (PILS) algorithm, a modified sensitivity encoding (MSENSE) algorithm, or a SPACE RIP algorithm.
This invention relates to medical imaging systems, specifically magnetic resonance imaging (MRI) systems, and addresses the challenge of improving image reconstruction quality and efficiency. The system includes a data acquisition module that collects raw MRI data from a subject, a reconstruction module that processes this data to generate images, and a control module that manages the reconstruction process. The reconstruction module employs at least two distinct reconstruction algorithms to process the acquired data. These algorithms may include a sum of squares (SOS) algorithm, a geometric average (GA) algorithm, a sensitivity encoding (SENSE) algorithm, a parallel imaging with localized sensitivities (PILS) algorithm, a modified sensitivity encoding (MSENSE) algorithm, or a SPACE RIP algorithm. The system dynamically selects or combines these algorithms based on the characteristics of the acquired data, such as noise levels, signal strength, or imaging parameters, to optimize image quality. The control module ensures that the reconstruction process adapts to varying conditions, enhancing the accuracy and clarity of the final MRI images. This approach improves the reliability of MRI diagnostics by reducing artifacts and enhancing detail in the reconstructed images.
11. The system of claim 9 , wherein to reconstruct an MR image, the processor further causes the system to: generate multiple coil images based on the MR data, wherein each coil image is generated using a part of the MR data received by a corresponding RF receiver coil of the multiple RF receiver coils; and generate the MR image based on the multiple coil images.
This invention relates to magnetic resonance imaging (MRI) systems and methods for reconstructing MR images from raw data acquired by multiple radiofrequency (RF) receiver coils. The problem addressed is improving image reconstruction efficiency and quality in MRI systems that use parallel imaging techniques, where data from multiple coils must be combined to form a final image. The system includes a processor configured to process MR data acquired by multiple RF receiver coils. To reconstruct an MR image, the processor generates multiple coil images, each derived from a portion of the MR data corresponding to a specific RF receiver coil. These coil images are then combined to produce the final MR image. The system may also perform additional processing steps, such as applying a sensitivity map to the coil images before combining them, or using compressed sensing techniques to enhance reconstruction speed and accuracy. The invention aims to optimize the reconstruction process by efficiently handling data from multiple coils, reducing artifacts, and improving image resolution. The techniques are particularly useful in clinical and research settings where fast and high-quality MR imaging is required.
12. The system of claim 11 , wherein to reconstruct the first MR image or the second MR image, the processor further causes the system to: for each point of a plurality of points in the imaged object, determine pixel coordinates of corresponding pixels in the multiple coil images relating to the point of the imaged object; obtain pixel values of the corresponding pixels in the multiple coil images of the point; and reconstruct the first MR image or the second MR image based on the pixel coordinates and the pixel values of the corresponding pixels in the multiple coil images of the plurality of points in the imaged object.
Magnetic resonance imaging (MRI) systems often use multiple receiver coils to capture signals from an imaged object, but combining these signals into a high-quality image can be challenging due to variations in coil sensitivity and noise. This invention addresses the problem by providing a method for reconstructing MR images from multiple coil images with improved accuracy and efficiency. The system processes signals from multiple coils to generate at least two MR images, such as a high-resolution image and a low-resolution image. For each point in the imaged object, the system determines the pixel coordinates of corresponding pixels in the multiple coil images and retrieves their pixel values. Using these coordinates and values, the system reconstructs the final MR image by combining the data from all coils. This approach ensures that the reconstructed image accurately represents the imaged object by accounting for spatial variations in coil sensitivity and noise. The method is particularly useful in applications requiring high-resolution imaging, such as medical diagnostics, where precise image reconstruction is critical. The system may also include additional features, such as adjusting image resolution or applying noise reduction techniques, to further enhance image quality.
13. The system of claim 9 , wherein the correction information includes a divided image, and to generate the corrected image, the processor further causes the system to: divide the first MR image by the divided image.
This invention relates to medical imaging systems, specifically magnetic resonance (MR) imaging, and addresses the challenge of correcting artifacts or distortions in MR images to improve diagnostic accuracy. The system processes a first MR image by applying correction information to generate a corrected image. The correction information includes a divided image, which is used to mathematically adjust the first MR image. The processor divides the first MR image by the divided image to produce the corrected image, effectively removing or reducing distortions present in the original MR image. This approach leverages division-based correction to enhance image clarity and accuracy, which is particularly useful in clinical settings where precise imaging is critical for diagnosis and treatment planning. The system may also include additional components, such as a display for visualizing the corrected image and a storage device for retaining the processed data. The method ensures that the corrected image is generated efficiently and accurately, improving the reliability of MR imaging results.
14. The system of claim 9 , wherein the correction information includes a divided image, and to generate the correction information, the processor further causes the system to: smooth the divided image to generate a smoothed divided image; and normalize the smoothed divided image to generate a normalized image.
This invention relates to image processing systems designed to correct distortions in captured images, particularly those caused by optical or environmental factors. The system addresses the challenge of improving image quality by generating correction information that compensates for such distortions. The correction information includes a divided image, which is processed through two key steps: smoothing and normalization. Smoothing the divided image reduces noise and artifacts, producing a smoothed divided image. This smoothed image is then normalized to generate a normalized image, ensuring consistent brightness and contrast across the corrected output. The system leverages these processed images to apply precise corrections, enhancing the accuracy and visual quality of the final image. The invention is particularly useful in applications requiring high-fidelity imaging, such as medical imaging, surveillance, and scientific research, where distortion correction is critical for reliable analysis. By integrating these smoothing and normalization steps, the system ensures that the correction information is both effective and adaptable to various imaging conditions.
15. The system of claim 14 , wherein to generate the corrected image, the processor further causes the system to: divide the first MR image by the normalized image.
The invention relates to medical imaging systems, specifically magnetic resonance (MR) imaging, addressing the challenge of correcting distortions or artifacts in MR images to improve diagnostic accuracy. The system processes a first MR image and a normalized image, which is a reference or baseline image, to generate a corrected image with reduced artifacts. The processor divides the first MR image by the normalized image to mathematically correct distortions, such as intensity variations or geometric inaccuracies, caused by factors like magnetic field inhomogeneities or patient movement. This division operation normalizes the first MR image against the reference, effectively removing systematic errors present in the original scan. The normalized image may be derived from a separate calibration scan, a pre-existing reference dataset, or an averaged image from multiple scans. The system may also include preprocessing steps, such as aligning the images or applying noise reduction filters, to ensure accurate division and correction. The corrected image is then output for further analysis or display, providing a more reliable representation of the scanned anatomy. This technique is particularly useful in clinical settings where precise image quality is critical for diagnosis and treatment planning.
16. The system of claim 9 , wherein the multiple RF receiver coils have different spatial sensitivities and receive MR signals in parallel.
A system for magnetic resonance imaging (MRI) includes multiple radiofrequency (RF) receiver coils with distinct spatial sensitivities, enabling parallel acquisition of MR signals. The coils are arranged to capture signals from different regions of an imaging volume simultaneously, improving imaging speed and resolution. Each coil is tuned to a specific frequency range and spatial sensitivity profile, allowing the system to reconstruct a high-fidelity image by combining the parallel data streams. The system may also include a controller to coordinate signal acquisition, a processor to process the received signals, and a display to visualize the reconstructed image. The parallel acquisition reduces scan time and enhances image quality by mitigating artifacts caused by motion or signal interference. This approach is particularly useful in clinical and research settings where rapid and detailed imaging is required. The system may further integrate with existing MRI hardware, such as gradient coils and main magnets, to provide a comprehensive imaging solution. The use of multiple coils with varying sensitivities ensures comprehensive coverage of the imaging volume, addressing limitations in conventional single-coil or uniform-coil designs. The system optimizes signal-to-noise ratio and spatial resolution, making it suitable for applications in medical diagnostics, material science, and biological research.
17. A non-transitory computer readable medium comprising executable instructions that, when executed by at least one processor, cause the at least one processor to effectuate a method comprising: receiving MR data of an object, the MR data being acquired by multiple RF receiver coils of an MR device; reconstructing, by the at least one processor, a first MR image based on the MR data according to a first reconstruction algorithm; reconstructing, by the at least one processor, a second MR image based on the MR data according to a second reconstruction algorithm, the second reconstruction algorithm being different with the first reconstruction algorithm; generating, by the at least one processor, correction information about the first MR image by dividing the first MR image by the second MR image; and generating, by the at least one processor, a corrected image with reduced inhomogeneity intensity based on the first MR image and the correction information about the first MR image.
This invention relates to magnetic resonance imaging (MRI) and addresses the problem of inhomogeneity artifacts in MR images caused by variations in radiofrequency (RF) coil sensitivity and other factors. The system uses a computer-readable medium with executable instructions to process MR data acquired by multiple RF receiver coils in an MRI device. The method involves reconstructing two MR images from the same raw data using different reconstruction algorithms. The first image is generated using a first algorithm, while the second image is generated using a second, distinct algorithm. The system then generates correction information by dividing the first image by the second image, which helps identify and quantify inhomogeneities. Finally, the system applies this correction information to the first image to produce a corrected image with reduced intensity inhomogeneities. This approach leverages dual-reconstruction and ratio-based correction to improve image uniformity without requiring additional hardware or complex calibration procedures. The technique is particularly useful in clinical and research settings where accurate image interpretation is critical.
18. The non-transitory computer readable medium of claim 17 , the first reconstruction algorithm or the second reconstruction algorithm including at least one of a sum of squares (SOS) algorithm, a geometric average (GA) algorithm, a sensitivity encoding (SENSE) algorithm, a parallel imaging with localized sensitivities (PILS) algorithm, a modified sensitivity encoding (MSENSE) algorithm, or a SPACE RIP algorithm.
This invention relates to medical imaging, specifically magnetic resonance imaging (MRI) systems that use parallel imaging techniques to reconstruct images from acquired data. The problem addressed is improving image reconstruction quality and efficiency in MRI systems that employ multiple receiver coils to capture data simultaneously. Traditional parallel imaging methods often suffer from artifacts, noise, or computational inefficiency, particularly when dealing with complex imaging scenarios. The invention provides a non-transitory computer-readable medium containing instructions for performing MRI image reconstruction using at least one of several advanced algorithms: sum of squares (SOS), geometric average (GA), sensitivity encoding (SENSE), parallel imaging with localized sensitivities (PILS), modified sensitivity encoding (MSENSE), or SPACE RIP. These algorithms enhance image quality by optimizing data processing from multiple receiver coils, reducing artifacts, and improving signal-to-noise ratio. The SOS algorithm combines coil signals by summing their squared magnitudes, while GA uses geometric averaging for noise reduction. SENSE and PILS improve spatial resolution by leveraging coil sensitivity profiles, and MSENSE refines SENSE for better accuracy. SPACE RIP further optimizes reconstruction by incorporating iterative techniques. The invention ensures robust and efficient MRI image reconstruction across various clinical applications.
19. The non-transitory computer readable medium of claim 17 , the method further comprising: generating multiple coil images based on the MR data, wherein each coil image is generated using a part of the MR data received by a corresponding RF receiver coil of the multiple RF receiver coils; and generating the MR image based on the multiple coil images.
This invention relates to magnetic resonance imaging (MRI) systems and methods for improving image reconstruction from multi-coil MRI data. The problem addressed is the need for efficient and accurate generation of high-quality MR images from data acquired by multiple radiofrequency (RF) receiver coils. Traditional MRI systems often struggle with noise, artifacts, and computational inefficiency when processing data from multiple coils, leading to suboptimal image quality. The invention provides a solution by generating multiple coil images, each derived from a portion of the MRI data received by a corresponding RF receiver coil. These coil images are then combined to produce a final MR image. The process involves reconstructing individual coil images from the raw MRI data, where each coil image corresponds to data acquired by a specific RF coil. By processing the data in this distributed manner, the system can reduce noise and artifacts while improving image resolution and clarity. The method leverages the spatial sensitivity of each coil to enhance image quality, particularly in regions where signal strength varies across the imaging volume. This approach is particularly useful in applications requiring high-resolution imaging, such as medical diagnostics, where accurate visualization of anatomical structures is critical. The invention improves upon prior art by optimizing the use of multi-coil data to achieve faster and more reliable image reconstruction.
20. The non-transitory computer readable medium of claim 19 , the method further comprising: for each point of a plurality of points in the imaged object, determining pixel coordinates of corresponding pixels in the multiple coil images relating to the point of the imaged object; obtaining pixel values of the corresponding pixels in the multiple coil images of the point; and reconstructing the first MR image or the second MR image based on the pixel coordinates and the pixel values of the corresponding pixels in the multiple coil images of the plurality of points in the imaged object.
Magnetic resonance imaging (MRI) systems often use multiple receiver coils to capture signals from an imaged object, but combining these signals into a high-quality image remains challenging. This invention addresses the problem of accurately reconstructing MRI images from multiple coil images by improving the process of mapping and combining pixel data from different coils. The method involves analyzing an imaged object by determining pixel coordinates for each of multiple points within the object across multiple coil images. For each point, the corresponding pixel values from the different coil images are obtained. The system then reconstructs the final MRI image by integrating these pixel coordinates and values from all coil images. This approach ensures that the reconstructed image accurately represents the imaged object by leveraging the spatial and intensity information from each coil's data. The technique is particularly useful in improving image clarity and reducing artifacts in MRI scans.
Unknown
March 24, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.